So
there were a lot of great writeups in the past few days about the photo
sharing sites, some coinciding with the upgrade release today by Picasa. 

John Battelle had the insightful piece Thoughts on Picasa and Google's Marketing Strategy.  Wired Magazine ran a nice comparison piece yesterday comparing the various photo sites out at present.

With
all of the recent attention given to Picasa, Flickr and other photo
sharing strategies, I thought the time was right to unveil, Photohawk,
the ultimate online photo resource.

Although Picasa and Flickr
have made good early entries into the online photo field, and Google
Image Search shows extraordinary promise, these sites still leave much
to be desired and the fictional Photohawk site takes image search to
the next level. So, what should the photo sites of today be doing to
truly capitalize on the immense upcoming opportunity in photo search,
sharing and storage?

1. Limits on bandwidth, storage, etc. as an
inducement to get people to upgrade to pay service might be an
interesting short term strategy but poor long term planning. The real
long term power and potential of the photo sharing experience is in
building the largest collection of desirable, sought after, viewed
images on the internet. To this end, the ideal photo site would offer
unlimited storage for all. The site would be supported by unobtrusive,
Google styled key word advertising. Upgrade packages would be sold and
offered allowing an individual to keep the advertising dollars
generated by clicks as an optional feature -- otherwise all ad revenue
would go to Photohawk as the trade off for providing you the unlimited
storage.

2. The problem with Google Image Search is in the
quality of the searches. Many searches are somewhat meaningless because
there are limited tools to refine search. Type in "Ween" and you get
15,200 images. Some of these images are obviously better than others.
Refine the search down to only "large' images and you still get 673
images. Now Ween the band is no Britney Spears but if I'm a Ween fan,
and I am, what I need is some kind of ranking system. This ranking
system should have three components: rank by photo views, rank by user
votes, rank by Photohawk Editors. The idea here is that there are
initially three different ways to judge the subjectivity of a photo.

A
photo view is perhaps the easiest to capture. User votes would be along
the lines of a "hot or not" concept where individuals could, as part of
an online photo community, rate and rank photos on a 1 to 10 scale
basis. The most difficult part would be the rank by Photohawk Editors.
Here the company would need to hire individuals who would basically
spend their entire day voting on photos in the archive. These would be
subjective ratings based on technical aspects of a photo (is it over or
under exposed, etc.) as well as content aspects.

At present Brandon Stone is running one of the hottest potential sites on the internet over at Photoblogs.org.
What Brandon has done is to incorporate a photoblog ratings system
which allows users the opportunity to select "favorite" photobloggers.
These bloggers end up on a ranking list and the self perpetuating
system keeps the top bloggers at the top. To counteract the self
perpetuating trend of the top bloggers receiving all of the traffic
simply because they are the top bloggers, Brandon also has another
ranking system for newcomers which quickly moves hot new talent up the
photoblogging chain.

The top photobloggers at photoblogs.org
truly do provide stunning work. It's very artistic much of the time and
technically vastly superior to what one might get from Google Image
Search.

In addition to user ratings on each and every individual
photo, Photohawk would hire editors, a la a Looksmart search model to
further refine the photo search process scoring photos individually and
allowing for searches that prioritize the very best images of what an
individual is looking for.

If I do a search for "Golden Gate
Bridge" on Google Image search I get 22,500 images of mediocrity. If I
do an image search or "Golden Gate Bridge" at Photohawk I may or may
not get 22,500 images but I will be able to see the most stunning
beautiful amazing images as selected by both the general public and
Photohawk editors which will be a far more satisfying experience.

My friend Tom Conrad over at Savage Beast is doing something similar with music by hiring music professionals to categorize, rank and rate music in different ways.

3.
There are an amazing amount of public domain artistic images available
but not easily found on the internet. Part of Photohawk would involve
teams of scanners and surfers that built large online libraries of
public domain artwork. There is no reason that someone should not be
able to do an image search for "Van Gogh" and get every single painting
that was ever painted by Van Gogh in perfect high res clarity.

This
service will open up art to the general public in ways never
experienced before and drive an incredible amount of traffic to the
advertising supported site. There are millions of public domain
paintings that at present are not cataloged in high res on the
internet. Recently I reviewed a plug in for the Microsoft Media Center
called Gallery Player.
This player attempts, for a fee, to offer artwork for your fancy
plasma. The service is very limited offering something like 30
paintings for sale at a buck a piece. Photohawk will have thousands of
images for free download with a simple "right click" "save as"
maneuver.

4. Tags. Tags are suddenly very hot and for good
reason. Navigating a too large library of images demands them. In
addition to general public tagging, Photohawk images would also be
supplemented by professional Photohawk taggers (editors) to further
refine the cataloging of online images. Photohawk would also develop an
MVP Award to hand out to volunteers based on their tagging and rating
work. Top prizes would be given out to the top MVPs (cheap labor).

5.
Geographic Tags. Alongside regular tagging would be geographic tagging
based on location. Travel is a huge business -- a huge, huge, online
business. The top ranked photos would be categorized by location even
down to the street address level. These photos would then be
incorporated into a mapping system that I could use for travel.

The World City Photo Archive
currently offers up about 10,500 photos of cities all over the world.
I'm glad to have some of my photos included in their library. Much of
the photos on this site though are accurate but not spectacular.

If
I knew that I were going to Prague, for instance, and could then pull
up the equivalent (through aggregation) of an amazing coffee photo
travel book of Prague and even click through to find the locations of
the sites for my upcoming visit, this would be a valuable resource. As
always, next to the brilliant full sized photos would be a small text
based advertising box.

Alternatively I could click on a section
of a map and top ranked photos would pop up providing a visual version
for the wi-fi savvy tourist.

6. Incorporating large image
libraries. One of the problems with all of the current photo sharing
sites is that they depend on uploading images which is a painfully slow
process. Photohawk would create a free service for "power users"
whereby they could send external hard drives in the mail and have all
of their images hosted at the site. They could then further synch the
online site and their hard drive to update their library going forward.

Although I can send 50 photos to Flickr, I actually have about
55,000 .jpg images at present. It would take months if not years to
upload all of these 55,000 photos to Flicker. And yet they could be
transferred via external hard drive in a day. Photohawk would recognize
that adding 55,000 images at once to the Photohawk library would be
worth the labor and expense of this kind of custom work for a power
user.

7. Photoblogging is hot. Creating the top photoblog
templates is not as easy. The format on the hottest photoblogs seem to
revolve around a single image on a page. These are custom templates
developed by users mostly in Movable Type and other more sophisticated
blogging packages. Photohawk would offer a number of templates that
individuals could use to build their photoblogs. This would encourage
the top photobloggers to use the site and upload their photos which
could be crawled, ranked and searched.

8. Although images could
be removed by the owner at will, all images would be hosted on image
search servers. This would alleviate the problem at present with Google
Image Search of so many of the photo results being bad or old links.
The trade off here, however, would be that Photohawk would have to be
diligent and responsive to removing copyrighted material when requested.

The
initial success of Photohawk would not be measured in subscription or
even advertising dollars. Rather, the success of the project would be
measured by the size of the library and the subjective quality of
search. These two factors would drive the traffic and make the site the
premier destination on the internet for image searches.